An ink jet recording apparatus is an image forming apparatus that uses one or more liquid drop jet heads as recording heads. An ink jet recording apparatus is used, for example, as a printer, a facsimile, a copier, or a multifunction copier having functions of a printer, facsimile, and copier. An ink jet recording apparatus jets drops of inks or recording liquids from its recording heads onto paper (not limited to a sheet of paper but also refers to any medium, such as an OHP sheet, on which an image can be formed using ink drops or liquid drops, and may also be called a recording medium, recording paper, recording sheet, recording material, medium, or the like), and thereby forms (records or prints) an image on the paper.
Such an ink jet recording apparatus is able to form, for example, four types (or halftone levels) of dots: no dot, a small-size dot, a medium-size dot, and a large-size dot. However, these four types of dots are not enough to reproduce a large number of gradation levels. Therefore, to reproduce a large number of gradation levels, dither methods are used. In a dither method, halftones are reproduced by using both density modulation (intensity modulation) and area coverage modulation.
In a binary dither method, the density value of a pixel at a coordinate point is compared with a corresponding threshold value in a dither matrix and binarized into 1 (printed or illuminated) or 0 (not printed or not illuminated) based on the result of the comparison. This method enables obtaining binarized data for area coverage modulation by just comparing the density values of pixels in an image data with threshold values in a dither matrix and therefore enables high-speed processing.
Also, there are dither methods that use three or more values. For example, when forming an image with an ink jet recording apparatus that can form dots in three sizes, three dither matrices are used and pixels are classified into 0 (no dot), 1 (small-size dot), 2 (medium-size dot), or 3 (large-size dot).
There are many types of dither matrices. For example, Bayer dither matrix, random dither matrix, and blue-noise dither matrix are well-known. These dither matrices are designed so that dots are not concentrated in an area but uniformly distributed in area coverage modulation, and therefore are called dispersion types. Also, there are dither matrices called concentration types that are designed so that dots are concentrated around a certain point. For example, there is a concentration type dither matrix in which submatrices are arranged to form a screen angle.
Patterns formed by concentrated dots are highly visible when printed and hide unevenness in image density caused by low paper quality or low printing accuracy. Therefore, concentration type dither matrices are widely used especially in commercial printing.
Patent document 1 discloses a method of producing a mask or a dither matrix made of threshold values with which pixels in multi-level image data are compared to convert the multi-level image data into a halftone image. This method includes steps of a) determining a halftone dot pattern for each of predetermined halftone levels and b) creating the mask with the halftone dot patterns obtained in step a). In step a), halftone dot patterns are determined independently for each halftone level.
Patent document 2 discloses a method of preparing a halftone processing mask used in a tone reproduction method in which, when converting a multi-tone image dot by dot into binarized or multi-level image data using a dither matrix, parts of the multi-tone image with a certain density are converted to form a line based pattern having a predetermined direction and other parts are converted to have high-pass filter characteristics. The halftone processing mask is designed so that the line based pattern includes dots that always synchronize with a recording sequence matrix of dots formed by a combination of multipassing and interlacing of a serial head.
[Patent document 1] Japanese Patent Application Publication No. 2003-046777
[Patent document 2] Japanese Patent Application Publication No. 2005-001221
Although concentration type dither matrices have advantages as described above, they have disadvantages too. For example, when forming an image using an ink jet recording apparatus, concentration of dots or concentration of ink in an area may cause bleeding or undermine fixation of the ink.
It is possible to design a concentration type dither matrix so that dots are not concentrated excessively. However, in highlight parts of an image, since only a small amount of ink is used, it is preferable to concentrate dots. On the other hand, in shadow parts, concentrated dots may cause problems. These two conflicting facts are making it difficult to design an appropriate dither matrix.
Also, on a coated glossy paper, ink does not spread smoothly and therefore it is comparatively difficult to evenly cover the surface with ink. On such a coated glossy paper, it is preferable to emphasize patterns using a concentration type dither matrix in parts with a low halftone level. However, using a concentration type dither matrix in parts with a middle or high halftone level may cause some dots to be left blank and reduce the image quality.
To reproduce smooth gradation using a dither matrix, the number of gradation levels of the dither matrix must be larger than that of input pixels. Generally, there are two methods to increase the number of gradation levels. A first method is to increase the size of a submatrix. A second method is to use multiple submatrices as a unit and thereby to increase the total number of dots. In the first method, when the number of gradation levels increases, the screen frequency decreases. Therefore, the first method is not suitable to produce a high-resolution image with a large number of gradation levels. The second method makes it possible to increase the number of gradation levels without decreasing the screen frequency. Since the dot areas (the numbers of dots generated) of submatrices are not always the same, if the ratio between submatrices with a large number of dots and submatrices with a small number of dots is unbalanced, the resulting pattern may look like a texture.
In a method where color planes of a color image, such as a CMYK image or an RGB image, are processed using a same dither mask, secondary or higher colors are formed by overlapping halftone patterns of different colors. When an image is formed using such a method on a paper on which ink does not spread smoothly, irregular white spots may be left in the image and, as a result, the image quality is reduced. Also, when an image is formed with a printing apparatus in which a dominant color is determined by the order of overlapping inks, such a method may cause hue distortion.